The THEMIS Constellation by wuyunqing


									                     The THEMIS Constellation
             P. HARVEY1, E. TAYLOR1, R. STERLING1 and M. CULLY2
     1. Space Sciences Laboratory, University of California at Berkeley, CA. U.S.A.
                      2. ATK Space Division, Beltsville, MD. U.S.A.
Abstract. The Time History of Events and Macroscale Interactions during Substorms
(THEMIS) mission is the fifth NASA Medium-class Explorer (MIDEX), launched on
February 17, 2007 to determine the trigger and large-scale evolution of substorms. The
mission employs five identical micro-probes (termed “probes”), which have orbit periods
of one, two and four days. Each of the Probes carries five instruments to measure
electric and magnetic fields as well as ions and electrons. Each probe weighs 134 kg
including 49kg of hydrazine fuel and measures approximately 0.8 x 0.8 x 1.0 meters (L x
W x H) and operates on an average power budget of 40 Watts. For launch, the Probes
were integrated to a Probe Carrier and separated via a launch vehicle provided
pyrotechnic signal. Attitude data are obtained from a sun sensor, inertial reference unit
and the instrument Fluxgate Magnetometer. Orbit and attitude control use a RCS system
having two radial and two axial thrusters for roll and thrust maneuvers. Its two fuel tanks
and pressurant system yield 960 meters/sec of delta-V, sufficient to allow Probe
replacement strategies. Command and telemetry communications use an S-band 5 Watt
transponder through a cylindrical omni antenna with a toroidal gain pattern. This paper
provides the key requirements of the probe, an overview of the probe design and how
they were integrated and tested. It includes considerations and lessons learned from the
experience of building NASA’s largest constellation.
Keywords: THEMIS, microsatellite, probe, constellation
PACS codes: 94.30.-d;; 94.30.cb;; 94.30.cj; 94.30.C-; 94.30.cp;
94.30.Lr; 94.30.Va; 94.30.Xy; 96.50.Fm.
   1. Introduction
    The THEMIS mission employs five simple, identical probes that fly independent and
synchronized orbits around earth. The orbit periods are designed to produce a combined
measurement set resulting from the apogee region conjunctions due to the natural evolution of
the orbits. The probes communicate independently with the operations center that operates
each probe in a round-robin serial fashion during normal operations. While the probes are
highly autonomous, attitude and orbit determination is maintained by the ground operations
center with all orbit and attitude maneuvers nominally taking place during ground contacts.
Thorough discussions of the mission design and instruments are presented in Vassilis et al
(2008) and the operations are described in Bester et al (2008). The ground-based
observatories measuring the aurora are described in Mende et al (2008).
    THEMIS was launched on a Delta II 2925-10 ELV from the Eastern Range into a direct
injection orbit near the final orbit of P3/4/5 (3 inner probes) in a stable 1.07 x 15.4 Re orbit.
The Probe Carrier (PC), a simple mechanical fixture bolted to the 3rd stage, dispensed the
probes, directly into this common initial orbit spin-stabilized at 16 +/-2 RPM. An on-board
hydrazine propulsion Reaction Control System (RCS) performed the final placement of each
probe into its final orbit with minor trimming prior to the prime science tail season.
    The spinning probes are passively stable, even under worst-case scenarios. The single-
string design is simplified by a minimal hardware complement, inherent functional
redundancy, with the instruments and the bus designed for graceful degradation. Analyses
demonstrated that either P3 or P4 can replace any other probe during the mission. Since four
probes can accomplish the minimum mission, THEMIS benefits from “constellation
redundancy” with a reliability of 80% for the nominal 2-year life and 93% for the minimum
1-year life.
    The five flight instruments (FGM, ESA, SST, SCM and EFI) are identical on all five
probes and were built using production methods. The instruments have adjustable data rates
to suit different orbit profiles and utilize heritage burst-data collection strategies incorporated
in the Instrument Data Processing Unit (IDPU), which has the single electrical interface to the

   2. Systems Engineering
    The engineering of a constellation of probe, rather than a single probe, influenced nearly
every discipline involved in the mission, from the number of probe, their level of redundancy,
their power and mass, their magnetic and surface charging cleanliness, to how to integrate
and test them, and even how to implement the NASA review process. From the project start,
engineers and managers understood they needed a Probe design which was simple to build,
test and operate. To implement “standard” probe features on multiple probe would likely
exceed the schedule and cost caps of the program.
    Redundancy. THEMIS developed the concept of “Constellation Redundancy,” namely,
having an on-orbit spare probe capable of replacing any other probe. This configuration has a
25% impact at most, while having fully redundant probes would have doubled the cost,
complexity and mass, none of which were practical alternatives in a cost-capped, single
launch vehicle approach.
    Probes use a basic single-string design, taking advantage of inherent redundancy and
having added redundancy only when mission critical and practical. Each probe has only one
Bus Avionics Unit (BAU) and one IDPU since it would be impractical to make these
redundant. On the other hand, the spin plane booms are inherently redundant with one another
every quarter rotation, and this is sufficient to meet the E-Field timing requirements. The

axial boom and magnetometer boom have redundant firing circuits because of mission
criticality and minimal increased complexity.
     Robust Design. Probes are both simple and robust as possible. Solar Arrays cover all six
sides and the BAU simply shed power loads automatically if the battery gets too low,
assuring a positive power configuration in any attitude. Probes, therefore, have no on-board
maneuver capability and their RCS cat bed heaters are nominally off. All maneuvers are
calculated in detail by Mission Operations, simulated on the ground and then uploaded to
each Probe as needed (Bester, 2008). In addition Probe attitude stability during on-orbit
deployments was extensively analyzed and was one fault tolerant in most cases (i.e Probe
deployment and EFI deployment). This architecture simplified the design, implementation
and cost of the Probes while keeping flight operators in complete control.
     Low-Power Design. Fitting multiple probes within the constraints of the launch vehicle
fairing limited the overall size of the probe solar arrays. This restricted available power to
almost all subsystems, except ones with relatively low duty cycle such as the transmitter.
Chief power users such as the flight computers have their clock frequencies as low as possible
to keep power to a minimum. The Catalyst Beds and Inertial Reference Units are powered
off until maneuvering. In order to survive earth eclipse of up to 3 hours, the surface materials
passively bias all temperatures up several degrees so that heater power is minimized in
     Separation Design. The design of the separation sequence and carrier changed several
times during the project, driven by both engineering and safety concerns. Initially, Probes
would be commanded autonomously by the individual Probe Bus Avionics Unit computer to
release from the carrier. A second concept incorporated a separation system timer with a
dedicated battery hosted on the PC to release the Probes. Both concepts were identified as
high-risk developments due to the nature of controlling explosives at the range and the critical
aspect of requiring the near simultaneous firing of the four lower Probes from the Probe
Carrier to avoid collisions. NASA management at GSFC and KSC recommended, and then
implemented, the extension of the launch vehicle third stage separation event onto the PC.
The extension lines were provided by the launch vehicle provider and qualified as if part of
the vehicle itself. Finally, at GSFC recommendations (based on previous mission lessons
learned), a Probe separation status system was added by ATK Space to diagnose if any probe
had not separated. This data was relayed to the ground through the launch vehicle telemetry
system and this, in fact, verified that all systems separated within 1 millisecond of the
commanded time.
     Reviews. THEMIS conducted a thorough review program with a NASA-provided review
team composed of GSFC and HQ-selected members called the Integrated Independent
Review Team (IIRT). As shown in Table 2-1, a total of 39 reviews were conducted at the
system and subsystem levels, 27 of which formally run by the IIRT. These reviews resulted
in 269 Requests for Action (RFA) and all actions were formally documented and closed by
the IIRT prior to launch.
     THEMIS benefited greatly by the experience and insights provided by the IIRT. Given
that the constellation presented new and unique challenges for the THEMIS team, the bi-
lateral discussions proved effective in improving the design and implementation of the
Review    No   Description
MSRR       1   Systems Requirements Review
IPDR       4   Preliminary Design Review (FGM, SCM, EFI/ESA/SST, GBO)
BPDR       1   Preliminary Design Review (Probe Bus & Carrier)
MPDR       1   Preliminary Design Review (Mission Level)
MCRR       1   Confirmation Readiness Review
MCR        1   Confirmation Review
ICDR       4   Critical Design Review (FGM, SCM, EFI/ESA/SST, GBO)

BCDR      1    Critical Design Review (Probe Bus & Carrier)
MCDR      1    Critical Design Review (Mission Level)
MOR       1    Operations Review
ITS       1    Instrument Test Summary
BTS       1    Bus Test Summary
IPER      3    Instrument Pre-Environmental Review (1, 2/3, 4/5)
IPSR      3    Instrument Pre-Ship Review (1, 2/3, 4/5)
BPSR      5    Bus Pre-Ship Review (1-5)
PCPSR     1    Probe Carrier Pre-Ship Review
MPER      2    Mission Pre-Environmental Review (Probe 2, 1/3/4/5)
MFOR      1    Mission Flight Operations Review
MPSR      2    Mission Pre-Ship Review, Delta Pre-Ship Review
MRR       1    Mission Readiness Review
LVRR      1    Launch Vehicle Readiness Review
FRR       1    Flight Readiness Review
LRR       1    Launch Readiness Review
Table 2-1. Formal Reviews

    Spares. Based upon past experience in the Cluster I&II programs, in which most of the
instruments relied upon their spares at one time or another, THEMIS chose to build at least
one spare of each instrument. This decision proved wise as three of the five instruments
swapped out sensors for the spare unit. Spare bus components included a battery, side solar
panel and top/bottom solar array although none of these were actually used.
    Radiation. THEMIS has a two-year design life, mainly driven by its radiation
environment and 100% total ionization dose margin. Originally planned for launch in Fall
2006 and having two winter campaigns in 2007 and 2008, the launch vehicle was delayed
until just past the winter campaign of 2007. To ensure that baseline objectives remain intact
and avoid a radical mission redesign late into the program, a 5 month coast-phase was
inserted into the mission giving it a total duration of 29 months, and cutting the radiation
margin to about 65%.

                 Figure 2-2. The Dose Depth Curve for Probe 3 and 4 Orbits

Figure 2-2 shows the approximate annual radiation dose encountered by the P3/P4
electronics inside increasing amounts of Aluminum shielding(Innovative Concepts, 2003).
The P1 and P2 orbits have less radiation exposure since they spend less time in the electron
belts, while P5 has a slightly higher dose rate based upon more time in these belts. Based
upon these simulations, THEMIS enveloped these radiation requirements and baselined using
5 mm of Aluminum (or equivalent) and 66 kRAD tolerant electronic parts. Together these
requirements were found to provide an achievable balance between parts costs and Probe
    Safety. Early in the development, the THEMIS team understood that satisfying range
safety would be an important goal. Having five fueled Probes, ten initiators and separation
explosive lines UCB and ATK Space began working the safety issues early to identify
hazards to personnel, flight equipment, and facilities. We developed appropriate design
interlocks and inserted safety steps into procedures sufficient to satisfy both NASA and the
Range. Instrument hazards included flight system high voltages and GSE liquid nitrogen
tanks. Probe and PCA hazards include RF radiation, lifts, transportation and fueling. UCB
and ATK Space tailored EWR 127-1, identified 20 Hazard Reports and delivered the Missile
Systems Pre-Launch Safety Package (MSPSP) on schedule to the Range. There were no
mishaps during the project.

   3. Payload
        The THEMIS science payload combined the science instruments into a single package
with shared data processing and storage capabilities. The instrument suite was designed, built,
tested and delivered as a single item for integration with each probe. While providing greater
scientific capabilities in on-board power and logic sharing, the approach also provided a
single electrical interface to the probe, allowed completely parallel instrument and bus
development schedules, while greatly simplifying probe I&T. The only exception to this rule
was the axial EFIs, which were contained in cylindrical composite tube that was integrated
with the Bus structure at ATK following test verification at UCB.
        The instrument complement consists of five instruments: the Fluxgate Magnetometer
(FGM), the Search Coil Magnetometer (SCM), the Electrostatic Analyzer (ESA), the Solid
State Telescope (SST) and the Electric Field Instrument (EFI). These sensors are controlled
by, and data is returned through the Instrument Data Processing Unit (IDPU) which has the
electrical interface to the Bus Avionics Unit. All harnessing from the sensors to the IDPU was
built and tested prior to Probe delivery, further simplifying subsequent I&T. Details of the
instruments are provided in REFS 1-6.
        Mechanical. Though none of the instruments have critical pointing requirements,
integration of the boom-mounted magnetometers nevertheless required precision
measurements both for balance calculations and for attitude determination (Pankow, 2008).
The FGM and SCM booms are one-shot deployment mechanisms responsible for holding the
sensors still with respect to the probe chassis. Thus, after the booms were mounted to the top
deck, precise measurements were made using a portable coordinate measuring machine
        The EFI provides 3D coverage once its Spin-Plane and Axial booms are deployed.
These boom systems are located on the Probe Center of Gravity (CG) so that the deployed
wires are orthogonal. The fuel tanks composite CG was aligned with EFI so that the wires
will stay orthogonal even as fuel is depleted. The boom deployment sequence had Spin Plane
Booms deploy first, followed by the Axial Booms in order to maintain spin stabilization.
Details of this sequence are given in Pankow (2008).
        The ESA and SST sensors poke through the corner panels in mid span. While the
SST heads were light enough to mount to the panel, the ESA was internally mounted to the
IDPU chassis for support at that elevation.
        Electrical. The IDPU to Probe electrical interfaces consist of a low speed
bidirectional serial interface for commands, housekeeping, and status information exchanged
between the Probe and instrument, as well as a high speed serial Clock and Data lines for
science telemetry. An 8 MHz clock and 1 Hz tick line combined with a Probe UTC message
provides synchronization of the two systems. The 8 MHz clock was used to synchronously
sample science quantities in the IDPU.
        The BAU provides instrument commands, time and probe status to the IDPU every
second using the serial interface. A buffered sun-sensor pulse is used by the IDPU for spin-
sectoring the SST and ESA data.
        The IDPU provides instrument housekeeping packets to the BAU, which is combined
with its data into CCSDS frames for downlink. Stored science data is transmitted over the
high-speed link when commanded to do so.
        Power. The Probe provides the IDPU a Main power service and an Actuator service.
The Actuator service is used for deployments.
        Thermal. Since each probe is very small, body mounted instruments were expected
to experience larger thermal extremes than in previous missions. Temperatures for the
instrument components were set at very wide ranges of -50C to +65C survival and -50C to
+50C operational.
    Contamination Requirements. While several THEMIS sensors are sensitive to
contamination, they were designed for easy handling and simple integration to the probe.
   The ESA and SST sensors are sensitive to molecular and particulate contamination at the
sub-micron level. Both have covers and an external purge provided by the instrument for
integration and test.

   The EFI sensors are sensitive to handling issues; i.e. asymmetries in the reflective
properties of the sphere, which would generate a spin-period photo-emission. Deployment
testing at I&T required a clean room environment and handling with gloves. The EFI also
required that all Probe surfaces be electro-statically clean to 10-8 Ohms/sqcm, which equated
to a requirement of limiting the total Probe exposed surface to have less than 1 cmsq of
insulating surface. All exterior surfaces and apertures, which are eclipsed had to meet this
    Typical sources of contamination on the Probe were easily mitigated to a satisfactory
level for THEMIS instruments. Wire harnesses, solar array panels, thermal blankets and
heaters were baked out prior to instrument I&T. The TV chamber was baked prior to probe
insertion, its contamination level monitored using a TQCM, during the test and backfilled
with GN2 at the end of TV.

Sensor    Key Requirement on Probe
FGM       Mag <1nT at 2m
SCM       Mag low AC fields
EFI       ESC <10e-5 ohms/cm2
ESA       Molecular < .01 ug/cm2
SST       Molecular < .1 ug/cm2

Fabrication and Test.
     Parts. Parts selection required GSFC-311-INST-001 and GSFC PPL-22, Appendix B de-
rating practices. In general, we used grade 3 parts with some up-screening to grade 2 for key
items in critical sub-systems such as the Bus Avionics Unit. We organized the “Common
Buy” program, purchasing parts for all instruments both for the raw cost efficiency as well as
to limit the number of different part types and purchasing lots needed. By limiting part types
and lots, we lowered the probability that we would have to open the Probes and replace parts
due to a NASA Alert.
     Manufacturing. The sheer number of subsystems drove manufacturing decisions
towards automated circuit board fabrication as well as internal test functions for each
subsystem. For example, the field instruments can stimulate all sensors and the particle
instruments can simulate counts and energies. We arranged for manual work to be performed
by the same technician for all units of the same design. This included hand soldering,
harnessing, thermal blankets and thermal taping. These actions yielded remarkably uniform
performance and substantially accelerated the flight test program.
     Flight Software. UCB developed the IDPU software using PC-based assembler and
linker products developed in-house. The 208 software requirements, specification and test
verification were actively reviewed by NASA IV&V, and their recommendations were very
helpful. For the most part, FSW performance analyses and data products used Excel
spreadsheets, and the modular software was tested on the IDPU engineering model. See
Taylor et al (2008).

Instrument Suite Integration.
        Instrument Integration and Test (II&T) was a two step process, in which sensors were
tested at the box level for unique functions, then integrated to the IDPU and flight harness
forming the Instrument Suite. This maximized the instrument level test time while
minimizing personnel and facility resources.
    For II&T, we used the UCB/SSL E125 clean room. Instrument harnessing was built
using mockups of the Probe deck, and Multi-Layer Insulation (MLI) blankets were made at
GSFC using instrument mockups. All harnessing and MLI were baked out in UCB/SSL
vacuum chambers.
   As expected, testing the first flight model (FM1) was pivotal in the maturation of
procedures for subsequent models. We tested subsequent models in pairs of instrument suites
subsequently; i.e. FM2,FM3 and FM4,FM5. The Instrument Ground Support Equipment
(IGSE) used a language nearly identical to the ITOS used by the probe bus. Thus procedures
which were developed at instrument I&T flowed with only minor modifications into Probe
I&T. This flow is shown in Figure 3-1 and a photo of a suite in Thermal Vacuum is shown in
Figure 3-2.

                        Harness Fab &      IDPU-Harness       Mag Boom           EFI Boom
     Platform Fab
                           Bakeout        Functional Test    Deployments        Deployments

  IDPU-EFI Functional     IDPU-SST          IDPU-ESA          IDPU-FGM           IDPU-SCM
         Test           Functional Test   Functional Test   Functional Test    Functional Test

    INST Functional       INST Self                                           THERMAL VACUUM
                                          EMC/EMI/MAG        VIBRATION
         Test            Compatibility                                            (6 cycles)

   Figure 3-1. Fabrication/Flow for Engineering and Flight Suites.                (Emissions and
susceptibility tests on ETU.)

   Figure 3-2. Instrument Suite in TV Preparation

   4. Probe Bus and Carrier
The THEMIS Probe is a highly optimized system that met the extreme challenges posed by
the mission. As illustrated in Figure 4-1, each probe consists of the bus subsystems and the
instrument suite, consisting of four EFI radial instruments, two EFI axial instruments, one
ESA, one pair of SSTs, one SCM, one FGM, and an IDPU. The bus subsystems include
Structural/Mechanical, Thermal, Reaction Control Subsystem, Attitude Control Subsystem,
Power, Communications, and Avionics.

                      Figure 4-1. Probe Bus and Instrument Subsystems

In order to implement the concept of “constellation redundancy,” each of the five probes is
identical in design and capable of being placed in any THEMIS orbit. The probe design was
driven by a number of requirements including
    • All five had to be small enough and light enough to be launched on a single launch
    • Assuming small solar arrays, each probe had to be power efficient;
    • Given the low mass, each probe had to use radiation-hardened electronics;
    • To implement the orbits, the design had to maximize its fuel carrying capacity;
    • To avoid contaminating the magnetic measurements, the design and components had
        to be non-magnetic and non-permeable;

   •    To reduce surface charges that would impact EFI, the exterior materials were
        conductive and grounded;
   • Given the 3-hour shadows while operating the instruments, the design included
        considerable thermal blanketing, thermostatically controlled heaters, and careful
        selection of surface materials;
   • In order to operate in any attitude/orientation, the structure design had to tolerate
        extreme temperature swings from -115C to +105C.
The major subsystem designs and how these subsystem designs achieve the mission
objectives are described in the following sections.

Structure and Mechanical Subsystem
        The THEMIS Probe Bus structure provides mechanical support for all other
subsystems and consists of ultra lightweight panels constructed of composite graphite epoxy
face sheets and an aluminum honeycomb core. All panels have embedded fittings of either
titanium and/or aluminum that have been machined to minimize mass.
        The probe is rectangular in shape with overall dimensions of appx 82x82x45cm in
order to provide simplicity and minimizing costs in the solar arrays. The structure is divided
into a lower deck, an upper deck, four corner and side panels. The lower deck is the primary
mounting surface for the instruments, propulsion system and Probe components. It also
interfaces to the probe separation system. The side panels double as substrates for the solar
cells. The exterior surface of the upper deck provides inserts for mounting the two
magnetometer booms.
        The instrument and Probe components are mounted to the lower deck simplifying the
load-bearing structure design and facilitating integration. The lower deck and separation
adapter fitting is the probe primary structure, carrying the load from all the internal
components, side panels and upper deck into the probe separation fitting and ultimately
through to the Probe Carrier and launch vehicle. The probe structure is designed so that all
four side panels could be removed during I&T to allow access to the internal components.
        The primary structure must also withstand the extreme temperature swings during
early orbit operations and eclipses. The design employed low conductance composite
structure and isolated solar panels in order to minimize internal thermal swings between full-
sun and shadow operations. Extensive analysis and development testing was performed on
the new composite elements of the structure. These environments are simulated via vibration
testing and panel level thermal cycling at the subsystem level prior to delivery of the probe
structure to Integration and Test (I&T). The mass of the entire structure and mechanical
subsystem including mounting hardware is 15 Kg and represents approximately 19.5% of the
Probe dry mass.

Thermal Control Subsystem
        The Probe thermal design was a challenge given the 3-hour eclipses, the need for
maneuverability and the probe’s low mass. Its thermal subsystem employs a hot-biased
design using solar heat to bias component temperatures upward so the probe can survive long
eclipses with minimal heater usage (less than 12 Watts orbit average). Additionally, the
design allows the probe to be thermally safe in nearly all sun aspect angles. Most of the bus
and instrument components mount directly to the honeycomb base plate and operate
nominally at ~30o C. Components are blanketed with Multi-Layer Insulation (MLI) and have
simple thermo-statically controlled film heaters. Thermistors are used for temperature
monitoring. High-efficiency MLI blankets minimize heat loss from the hydrazine Reaction
Control System, which must always remain above 5°C to keep the fuel from freezing in the
lines. The probe includes external coatings with high solar absorptance-to-emittance ratios,
such as Vapor Deposited Gold (VDG). In order to reject the transponder heat, Optical Solar
Reflectors (OSR’s) are used on the bottom of the probe.

Reaction Control Subsystem (Propulsion)
         Each THEMIS Probe includes a Reaction Control Subsystem (RCS) to correct launch
vehicle dispersion errors, inject each probe into its respective mission orbit, maintain the
orbits, adjust spin-axis pointing and maintain a nominal spin rate. The fundamental
robustness of the mission design is due to the capability of probes 3 or 4 (P3, P4) to fully
replace any probe, should it fail. Thus, the RCS has been sized for a nominal mission profile
plus the worst-case contingency of replacing the P1 probe.
         The probe is capable of both axial and side thrusting for orbit maneuvers with
minimal efficiency loss allowing for operational flexibility. The tangential thrusters also act
individually for spin rate adjustment. The system consists of two fuel tanks, four 4.4 Newton
thrusters, a pressurant tank, latch valves, pyro valves, and miscellaneous hardware.
         The two lightweight fuel tanks hold up to 49 kg of hydrazine (in total) and were
specially designed and qualified for the THEMIS program. The tanks are made of high
strength alloy (inconel) and are supported by the bottom and top panels via integral polar
fittings. Tanks were verified non-magnetic by testing at UCLA. A high-pressure Carbon
Overwrapped Pressure Vessel (COPV) tank and pyrotechnic actuated valve dramatically
enhance the systems capability. Once the fuel in the tanks has been depleted by
approximately 25%, ground personnel command the pyrotechnic valve to open which
connects the high pressure tank to the fuel tanks. The resulting increase in pressure provides
significantly more delta-V, totaling 960 meters per second.
         The two axial engines provide 4.4 Newtons of thrust allowing for major orbit changes
of the probe. In addition, two tangential engines of the same size provide either spin control
or lateral thrust to the probe.
         In order to maintain mass balance throughout the mission life, the two tanks were
mechanically arranged to allow for symmetric fuel depletion. The pressurant and propellant
sides of the RCS are interconnected to provide both symmetry and added probe reliability.
Latch valves are located strategically to prevent adverse propellant migration during the
launch phase of the mission. Following launch, both latch valves were opened to take
advantage of stabilizing propellant migration inherent in this configuration.
         Tank, line and thruster heaters are thermostatically controlled to maintain the
hydrazine propellant comfortably above its freezing point. Thruster catalyst bed heaters are
controlled by the BAU. The Flight Operations Team preheats the catalyst bed 30 minutes
prior to firing in order to prevent cold-start degradation. The entire RCS weighs only 12 kg
without fuel and is approximately 15% of the Probe dry mass.
Attitude Control Subsystem
         The Attitude Control Subsystem (ACS) measures sun pulses and vehicle motions
needed to support maneuvers, spin rate control and science data analysis. The ACS
components are the Miniature Spinning Sun Sensor (MSSS) and the Inertial Reference Unit
(IRU). The MSSS provides the sun elevation once per spin and assists in the calculation of
spin rate. Using multiple spin pulses, flight software is able to determine the spin rate. The
IRU is a solid-state assembly which measures angular rate of motion of the probe in X and Y
axes. While these devices provide probe-relative data, the near Earth FGM data are used to
verify probe absolute attitude once per orbit.
         ACS telemetry is transmitted to the ground where it is processed into physical
coordinates. If maneuvers are required, ground systems calculate the necessary commands to
be sent to the probe, and these commands are verified on a ground-based avionics simulator
prior to application in space. The ACS Bus components together weigh only 0.6 Kg. Figures
4-2 and 4-3 show the Miniature Sun Sensor and IRU assembly.
         While extremely simple, the ACS design depends completely upon both the spin
stability of the probes throughout all mission phases and the knowledge of the FGM sensor
with respect to the probe body. To achieve spin stabilization, the probes are configured to
have their center of mass closely aligned to the geometric axis. This alignment is
accomplished through painstaking placement of components and by adjusting balance masses
prior to launch. Careful design and measurements of the FGM boom, its repeatability and
stiffness in thermal extremes were essential in providing accurate attitude knowledge to
mission operations.

Figure 4-2 Miniature Sun Sensor

Figure 4-3 IRU Assembly

Power Subsystem
        The Power subsystem is designed to provide all of the necessary power for the bus
and instrument subsystems for the life of the mission in both sunlight and during eclipse. The
power system is a Direct Energy Transfer (DET) system with the battery and solar array
connected directly to the power bus. The solar array power control and battery charging are
performed using linear and sequential switching shunts.
        Each probe has eight solar arrays that provide power generation in any orientation.
There are two arrays mounted on the bottom and top decks and there are four side panels. The
arrays use high efficiency cells that are bonded to the composite substrates. The side panels
are also primary structure that adds to their design complexity since they have to transfer
loads between the top and bottom decks. In order to reduce surface charging, all the cover
glass is electrically grounded to a common ground on each panel. This is accomplished by
bonding a highly conductive grid onto the panels following cell placement.
         The probe is highly efficient in power usage with approximately 36.85 Watts required
in full science mode for a 24-hour orbit, which includes a 3-hour eclipse and a 30-minute
transmitter turn-on. The capability for that orbit at the mission End of Life (EOL) is 40.35W.
The top and bottom panels are intended to provide approximately 20W at EOL, which is
sufficient power to enable the probe bus components to survive anomalous attitudes in a low
power condition.
         High efficiency triple-junction Gallium Arsenide solar cells are used. The cells are
approximately 4 x 6 cm, and they have an average BOL efficiency of 27% at room
temperature. Each side panel has four strings and each top and bottom panel has two strings.
Each string consists of 20 series cells with integral bypass diodes. Strings are carefully
arranged on the panels to cancel the magnetic field generated by cells. Cover glasses are 8 mil
thick with UV reflective coating. The cover glasses also have ITO coating to provide
electrical conductivity and electrostatic cleanliness. The cover glasses are inter-connected by
conductive epoxy to provide a bleed-path for surface charges to chassis.
         Power is stored onboard by a Lithium-ion battery that maintains full probe power
during eclipse. The battery is lightweight and has a 12.0 Amp-hour capacity. The major
power subsystem components weigh approximately 10.3 kg and represent approximately 13%
of the total Probe bus dry mass.

Communication Subsystem
        The Communication subsystem consists of an S-Band transponder, diplexer and
circularly polarized antenna mounted to the center boom structure.
        The antenna consists of six receiver/transmit stack patch antennas and a power
divider. These antennas are extremely lightweight and must have a conductive surface in
order not to build up surface charge. For high data rate communication, they provide
increased gain in a ±45° band about the plane perpendicular to the spin axis. Although
reliable communication is possible outside this region, there is a small null region about the
antenna boresight, and if the probe orientation is such that the line-of-sight to ground falls
within this region, communication is not possible. However, since the probe is inertially
pointed, communication outages of this kind would last for only small fraction of an orbit.
        The antenna is connected to the transponder via the diplexer. The CXS-610
transponder contains a receiver and a 5W transmitter. The antenna is always coupled to the
receiver with no switches in the receive path, and the receiver is always powered. It
demodulates command signals and outputs both data and timing to the BAU. For telemetry,
the BAU provides baseband signals to the transmitter which phase modulates them onto the
carrier. The transponder can be operated in a coherent mode that provides turn-around
ranging capability.
        Robust link margins exist for the uplink, even for the case of Probe 1 at apogee. For
the downlink, multiple telemetry rates ranging from 1 kbps to 1024 kbps are provided. The
total mass of the communication subsystem is 3.2 kg and represents 4% of the Probe dry
mass. Figures 4-4 shows the transponder. The S-Band flight antenna is shown in Figure 4-5.

Figure 4-4 S-Band Transponder

Figure 4-5 S-Band Antenna

         The Bus Avionics Unit (BAU) provides numerous functions for the probe bus and
contains the flight computer. The BAU provides the communication interface, instrument
electrical interface, data processing and power control for the probe bus. It contains five
stacked modules with a total weight of 3.0 kg and average power of 7 Watts. See Figure 4-6.
         The Data Processor Module (DPM) contains a radiation-hardened computer featuring
a Cold Fire processor operating at 16 MHz. This module performs all the onboard processing
and data handling using 64 MB of bulk memory and a 2.1 Mbps data interface with the
instrument electronics. The BAU hosts the RTEMS real-time operating system and the
application control and data handling software for the probe bus. Instrument and bus
housekeeping data is stored in the local bus memory with science data stored in the IDPU.
During a ground pass the housekeeping data is transmitted directly by the processor, while
science data is copied out of the IDPU to the transmitter.
         The Communications Interface Module (CIM) receives the command bit stream from
the receiver and provides CCSDS blocks to the processor. The card also processes a limited
number of hardware commands that may be received from the ground and executed without
the intervention of the processor. The card provides downlink telemetry data to the
transmitter for transmission to the ground. The data may be real-time engineering, playback
engineering data, and playback science data from the IDPU. Multiple telemetry rates are
provided, ranging from 1 kbps to 1024 kbps. All data are encoded with rate 1/2 convolutional
and Reed-Solomon encoding. The communications card also provides a hardline telemetry
data stream for ground testing.
         The Power Control Electronics (PCE) has 3 modules needed to control the solar array
shunts, regulate battery charge, generate and distribute secondary voltages, generate and
distribute discrete commands, and monitor separation signals from the launch vehicle and
initiate probe separation from the probe carrier. The PCE also contains circuitry needed to
condition and digitize most of the analog signals on the probe including IRU rate signals and
temperature sensors.
         The battery is charged at a fixed rate until the battery voltage reaches a commandable
limit, at which point the charge current switches to trickle charge. The upper voltage limit
can be selected conservatively so that no cell balancing is required. The BAU has the ability
to autonomously remove power from the IDPU in case of over-current or battery under-
         The BAU receives uplink commands at a fixed rate of 1000 bps. Commands are
received using CCSDS protocols that guarantee correct, in-sequence delivery of variable-
length command packets. All command transfer frames undergo several authentication
checks. Invalid frames are rejected and the rejection is reported in telemetry. Commands
sent to the probe will either be executed in real time or stored for later execution. Two kinds
of stored commands are provided: Absolute Time Sequence (ATS) commands and Relative
Time Sequence (RTS) commands. ATS commands have time tags expressed in UTC times,
with a resolution of 1 second, specifying an absolute time of day. RTSs are command
sequences that include relative delays between commands.
         Since THEMIS orbits have long periods between contacts as well as radiation belt
exposures, the BAU provides 64 MB of engineering data storage with error detection and
correction (EDAC). Playback data stored in bulk memory is formatted into multiple
segments, called virtual recorders, which allows for easy segregation into different types of
data such as bus engineering data, instrument engineering data, event files, etc. The size of
the virtual recorders is adjustable, allowing the memory to be remapped in the event of failed
locations. The integrity of the data stored in bulk memory is maintained by memory
scrubbing software, which uses the EDAC to correct single bit errors. Operating at a low
priority, the memory scrub task cycles through all the data stored in bulk memory once per
         The BAU maintains a precision onboard clock and distributes time to the IDPU in
Universal Time Code (UTC). Time synchronization between the bus and the payload is
achieved by the use of synchronous 8 MHz and 1 Hz clock signals sent to the IDPU. The
IDPU uses the 8 MHz to collect science data. Once per second, the BAU sends the IDPU the
UTC that will be valid at the next 1-Hz pulse. Together these actions allow the bus and
instrument to properly time-tag all science and engineering data.
         The BAU provides several autonomous functions that insure the health and safety of
the probe while out of ground contact. A watchdog timer is provided which continuously
monitors processor operations, and should a processor malfunction be detected, restarts the
processor automatically. A checksum routine runs at low priority, checking static areas of
memory. A telemetry and statistics monitoring function is provided which performs "limit
check" operations on the data and which maintains telemetry statistics. If pre-specified
conditions occur, it can initiate the execution of a stored command sequence.
         The BAU utilizes system tables to implement operational controls and to ease ground
system operations. System table operations constitute the primary ground interface for probe
control functions such as stored command operations and modifications of on-board
parameters. The BAU also has the ability to build "memory dwell" packets to monitor any
memory location for diagnostic support.

Figure 4-6 Flight BAU

Probe Carrier and Separation System
         The THEMIS Probe Carrier and Separation Systems met the programmatic challenge
to launch all five probes on a single Delta II. The engineering challenges included imparting
an initial stabilizing spin rate to each probe, maintaining a positive separation between probes
and doing so even if one probe failed to separate.
         The launch vehicle design had the Probe Carrier bolted to the third stage of the Delta
II. At the end of the launch sequence, the third stage despun to 16 RPM and initiated
separation pyrotechnics. As designed, the top probe deployed first and the lower four probes
deployed simultaneously three seconds later. Launch vehicle analog telemetry confirmed the
correct release profile and probe telemetry confirmed expected probe motions. Initial spin
rates for all five were between 16 and 17 RPM while sun angles were within 6 degrees of
each other.
         There were a number of significant challenges in the separation system design. First,
since the Probe Carrier was spinning at release, the four lower separation systems had to
operate reliably with a side load. Second, in order to avoid collisions between probes and/or
the carrier, the separation system had to move the probe quickly away from the separation
plane while imparting a low tip off rate.
         The Probe Carrier is predominately aluminum alloy, is weight optimized, and includes
a patch panel that manifolds all of the umbilical electrical and control circuit cabling from the
probes to the launch vehicle. The separation system was extensively analyzed and tested to
properly characterize its performance and to verify all of the mechanical parameters that drive
the overall Probe and Probe Carrier system clearance verification analysis. Figure 4-7 shows
the Probe Carrier at Astrotech Space Organization (ASO) launch site payload processing

Figure 4-7. Probe Carrier Ready for Probes

                             Observatory Facts
Number of Probes        • Five
                        • Probe Bus Dry Mass: 51 kg
                        • Instrument Mass: 26 kg
                        • Probe Dry Mass: 77 kg
                        • Propellant: 49 kg
                        • Probe Wet Mass: 126 kg
                        • Allowable Mass: 134 kg
                        • Probe Bus Power: 11 W
                        • Instrument Power: 15 W
                        • Heater Power (EOL/24 hr orbit/3 hr eclipse): 11 W
                        • Probe Power: 37 W
                        • Available Power: 40.5 W
                        • Battery capacity (BOL): 12 AHr
                        • S band
                        • EIRP: 2.4 dBW
Communications          • Two-way Doppler tracking
                        • Uplink command rate: 1 kbps
                        • Downlink telemetry rates: 1 kbps to 1.024 Mbps
                        • Command and telemetry format: CCSDS Version 1
C&DH                    • Engineering data storage: 64 MB, 5 days worth
                        • Timing : 8 MHz, 1 Hz and UTC distribution
                        • Spin rate (Science): 20 rpm
                        • Spin axis orientation: < 1° (knowledge), < 3° (control)
                        • Spin phase knowledge: < 0.1°
                        • Ground based attitude determination
                        • Monopropellant Hydrazine System
                        • Number of thrusters: 4 (4.4N ea.)
                        • Total ΔV: 940 m/s
                        • Propellant: 49 kg
                        • Probe Carrier Mass: 147 kg
Probe Carrier           • Total Payload Mass: 777 kg
                        • Mass to orbit capability: 829 kg
                        • Instruments
                         o Flux Gate Magnetometer
                         o Search Coil Magnetometer
                         o Electrostatic Analyzer
                         o Solid State Telescope (x2)
                         o Electric Field Instrument Radial (x4) and Axial (x2)
Science Instruments
                        • Booms
                         o 5-m axial booms (x2)
                         o 20-m radial booms (x4)
                         o 1-m SCM boom
                         o 2-m FGM boom
                        • Instrument Data Processing Unit
                        • Data Volume: ~ 400 Mbits per day
Science Data Volume
                        • 5 days worth of storage
Radiation Environment   • Total dose: 66 krads (2 years, 5mm Al shielding, RDM of 2)
                        • Observatory Ps = 0.91 (2 years)
                        • Mission Ps = 0.94 (4 of 5 s/c required for mission success)

5.     Integration and Test
5.1    Probe Integration at UCB
     Integration of the Probe buses and Instrument Suites was performed at UCB/SSL in a
class 10000 clean room. Each instrument suite, complete with flight harnessing, was rolled
up to a Probe and electrically connected via extension cable. Following interface verification,
all instrument components were mechanically integrated and alignments verified. Due to the
proximity of the Berkeley Ground Station (BGS) and Mission Operations Center (MOC), end
to end verification of RF communications with BGS and MOC were verified at this point.

5.2    Environmental Verification Testing
   Environmental testing of the THEMIS probes was conducted at the Jet Propulsion
Laboratory (JPL) in Pasadena, California, in two stages: first with a single “pathfinder” Probe
F2 and later with all five Probes. While other probe buses and instrument suites were still in
subsystem test, F2 proceeded to JPL and through environmental testing. The purpose was to
uncover problems before fully integrating the other Probes.

5.2.1 Pathfinder Environmental Tests
    The Pathfinder schedule and test sequence is shown in Figure 5.2.1-1. The test team
included staff from UCB, ATK Space and JPL. Daily teleconferences were held to coordinate
staff and plans, and to discuss and resolves questions and issues as they arose. During the
pathfinder test, we verified the mechanical ground support equipment used for lifting, rotating
and manipulating the probe. We also tested and verified the electrical ground support
equipment used to monitor and command the probe.
                DATE           ACTIVITY
                Mar 20-21      Arrival, offload and set up.
                Mar 22         Magnetics Testing
                Mar 23-24      Move to Vibration facility and set up.
                Mar 27-30      Vibration Tests X, Y & Z axes
                Mar 31         Separation Shock Test
                Apr 3          Move to EMC facility and set up.
                Apr 4-6        EMC Tests
                Apr 7          Move to Thermal Vacuum Facility and set up.
                Apr 10-11      Thermal Vacuum closeouts and blanketing
                Apr 12         Install probe in TV chamber
                Apr 13-16      Thermal Balance
                April 17-21    Thermal Vacuum Cycles (4x)
                Apr 24         Move to Magnetics Facility, Final Mag Test
                Apr 25         Transport to UCB/SSL.
                    Figure 5.2.1-1 Probe F2 Environmental Test Sequence
    The magnetics survey took place late at night to avoid interference from vehicles passing
outside the building. Three axis magnetic field measurements were taken: [1] all instruments
off; [2] all instruments turned on; and [3] after 15 Gauss de-perm. The Probe had
measurements taken in both vertical and horizontal positions, with several rotations in each
position. Measurements showed the Probe to be well within the specified requirement of 5 nT
at 2.5meters from the center.

  Figure 5-xx. Probe F2 in EMC

Figure 5-xx. Probe F2 in Magnetics

                                Figure 5-xx. Probe F2 in Vibration

5.2.2 Full Payload Environmental Tests
    While the pathfinder was in environmental test at JPL the other four probe buses and
instrument suites were in assembly and test at ATK Space and UCB. The final buses were
delivered in May and June, enabling completion of Probes 3, 4, 1 and 5, a combined Pre-
Environmental Review and return to JPL by the end of June.
    Figure 5.2.2-2 summarizes the activities of the five Probes and carrier in environmental
testing. After arriving at JPL, the Probe Carrier was tested with the probe mass dummies,
proving that the carrier was capable of sustaining the worst-case masses of the fueled probes.
After probe post-ship functional tests, the Probes were integrated to the Carrier while still on
the vibration table (see Figure 5.2.2-1). Environmental testing of the five probes was
completed on schedule with no major anomalies.
Week      Major Tests and Activities
July 10   Post-Ship Probe Functional Tests
          Probe Tank Pressurization
          PCA Integration
          PCA X and Y axis Vibration
July 17   PCA Z axis Vibration
          PCA Acoustics Test
          Probe Separation-Shock (5x)
          Probe Tank Depressurization
July 24   F3, F4 Thermal Vacuum Set Up             F1, F5 Magnetics Survey
          F3, F4 Thermal Balance                   F1, F5 Spin Balance
          F3, F4 Thermal Vacuum Cycles (4x)        F1, F5 Alignment Measurements
July 31   F2, F3, F4 Magnetics survey              F1, F5 Thermal Vacuum Set Up
                                                   F1, F5 Thermal Balance
Aug 7     F2, F3, F4 Spin Balance                  F1, F5 Thermal Vacuum Cycles (4x)
          F2, F3, F4 Alignment Measurements        F1, F5 Move to Bldg 179

Aug 14    F2, F3, F4 Functional Tests              F1, F5 Spin Balance (repeat)
                                           F1, F5 Functional Tests
            Figure 5.2.1-2 Full Payload Environmental Test Sequence

    Figure 5.2.2-1. Integrating the Probe Carrier Assembly on Vibration Table

              Figure 5.2.2-2. Probe Carrier Assembly ready for Z-axis vibration

               Figure 5.2.2-3. Probes 3 and 4 (Enclosed) at Thermal Vacuum
5.3    Launch Site Activities
    Following a four-month delay due to launch vehicle issues, the project received the green
light to proceed to launch in December. Initial pre-launch activities took place at Astrotech
Space Organization (ASO) located very near KSC. The flow of activities is shown in the
chart below.

    from JPL
                  DEC 11                                                                                                                          DEC 21
                                                                      Install         Battery                              Pressure/Leak
                      5 Probes                            Weigh
                                       Unpack, Setup                 Clamp ,          Charge,                               Test of RCS               Move Probes to
                 plus GSE ARRIVE                         Clamp,                                           LPT
                                          EGSE                         Bolt          Aliveness                                (Leave                   HPF / Bldg 2
                   at ASO Bldg 1                         harness
                                                                     Cutters           Tests                                Pressurized)
Probe Carrier
  Arrives at
 ASO Bldg 2

PCA Bag Fit
                                                                      JAN 3
                            Setup                                                                                                        Probe                        Weigh
                                          Weigh          Move        Set-Up                               Weigh           Probe/
                           EGSE,                                                         Fuel                                          Closeouts/                    Probes in
                                           Dry         Probes to     Fueling                               Wet              PCA
                          Aliveness                                                     Probes                                         Red/Green                       Flight
Side-by-Side                              Probes        Stands       Manifold                             Probes         Aliveness
                            Tests                                                                                                         Tags                        Config
Test PC with
  3rd Stage

 Install Sep
Harness, PC
                                                       JAN 15                                                                                              JAN 31
    PROBE                                                Spin                                                            Integrate
                    Integrated           PCA                           PCA                Weigh            PCA                                                Transport to
   CARRIER                                             Balance                                                            with 3rd         Canister
                     on Probe         Functionals                   Functionals           PCA             Bagging                                                 Pad
  Install SSS                                           (Wet)                                                              Stage

                                          FEB 1                                                  FEB 7                                                                   FEB 15

 Mate Payload/                                                                                                       Launch
                      Safe to          Probe/PCA         RF Chirp         Mission            Purge Bag,
  3rd stage to                                                                                                      Readiness
                       Mate            Functionals        Tests          Rehearsal             Install
      Delta                                                                                                          Review

                                            Figure 5.3-1 Payload Processing at ASO/KSC

As illustrated in Figures 5.3-1 to 5.3-5, important activities were as follows:
   • The five probes and carrier arrived at Astrotech on December 11, 2006.
   • Functional tests, pressurization tests, and bolt cutter installation were carried out in the
        Payload Processing Facility (ASO1);
   • The Probe Carrier was electrically checked in a “side-by-side” test with the 3rd stage,
        and integrated with its Separation System pyrotechnic lines in the Hazardous
        Processing Facility (ASO2);
   • Following the holiday break, Probes were moved to ASO2. Probes were weighed
        individually, fueled all at once, then re-weighed individually;
   • The probes were integrated to the PCA, electrically checked and spin-balanced;
   • On January 29, the PCA was bagged, lifted and mated with the Boeing 3rd stage;
   • On February 3, the containerized 3rd stage and payload were transported to the pad;
   • At the pad, the THEMIS team tested and charged probe batteries and practiced launch
        sequences while the Boeing Delta II rocket was being prepared for launch;
   • After delays due to lightning and high altitude winds, THEMIS was launched on
        February 17, 2007.

 Figure 5.3-1. Fueling probe in Hazardous Operations Facility

Figure 5.3-2. Probe Carrier Assembly on Spin Balance Machine

Figure 5.3-3. PCA Mate to the 3rd Stage

Figure 5.3-4. Delta II Launch Vehicle

Figure 5.3-5. Launch, February 17, 2007.

   6. Lessons Learned
The development of the Constellation proved demanding on several levels. The following are
conclusions from the experience:
   • THEMIS validated the effectiveness of the “pathfinder” approach, and showed
       dramatic improvement in performance, schedule, and cost of subsequent units;
   • Keeping the Probes identical, despite the fact that the mission required the Probes to
       be in different orbits, greatly benefited the test sequence;
   • Using the same technician for similar tasks across all Probes proved effective in
       maintaining similar Probe performance;
   • Requiring the instruments be designed with internal test features limited the need for
       drag-on test equipment through I&T;
   • Vibrating the full PCA rather than individual probes (or pairs of probes) was a
       significant improvement. It made the test more flight like, saved schedule and
       provided useful experience with the full PCA before heading to the launch site;
   • Following the Pathfinder Probe#2, performing the thermal vacuum test of the
       subsequent probes in sets of two worked well and saved considerable schedule time
       and costs;
   • Enclosing the probes within individually controlled thermal enclosures added
       substantial work and complexity. Despite this, the conclusion of the thermal engineers
       was that it resulted in a significantly better test and resultant thermal modifications.
   • Environmentally testing the pathfinder was very useful and resulted in numerous
       improvements, especially in thermal design. Thermal blanket modifications to the
       transmitter radiator were suggested by the pathfinder tests. These modifications
       significantly improved transmit durations and were verified in subsequent thermal
       vacuum testing on other probes.
   • Probe spin balance resulted in all five being balanced within specification with
       approximately 1.6 kgs total balance mass. Substantial benefits were realized with the
       number of probes. The balancing of the first two probes was time consuming and
       initially indicated need for substantially more balance mass than expected (over 3
       kgs.). Corrections made with the later probes showed that balance could be improved
       and the balance mass reduced. The two probes that were the first to be balanced were
       then re-balanced. The five probes were shown to have very high degree of uniformity
       and almost identical balance mass.
   • The four sets of electrical GSE proved to be uniform and we could operate any Probe
       from any GSE. This fact greatly facilitated the schedule as the GSE often stayed in
       one spot while the Probes moved through the facilities and were attached to any GSE
       test set.

    Problems in Integration and Test were concentrated on the first unit and fell off with each
pair of subsystems tested. As the instrument suites and Probes were tested, Problem/Failure
Reports (PFR) specified all units needing modification. For design errors or common
fabrication errors, all units were modified, including units that had not yet been tested. Of the
171 Bus and Instrument I&T PFR’s, 63 corrected multiple units providing down stream
benefits. While modifications proceeded on the current unit, changes of future units occurred
in parallel, effectively advancing the schedule of future units. As shown in Table 6-1, the first
unit found more than half the problems, followed by a rapid drop in problems in later units.
Since the last unit was on the critical path, this had a real effect on the mission level schedule.

                   I&T New PFRs        1st       2nd    3rd       4th       5th       Total
                   Probe Bus in I&T      35         3         2         1         6     47
                   Inst Suite in I&T         9      3         3         3         3     21
                   Inst Suite            48        19     20            9         7    103
                   Total                 92        25     25        13        16       171

                  Table 6-1. Problems Encountered at I&T (by Unit Order)

   7. Project Performance
7.1 Technical Performance
    The Probes and Carrier performed very well through integration and test, components
accumulating an average of 810 hours overall and 250 hours in thermal vacuum conditions.
Probes had an average of 350 failure free hours at launch. Of the 476 mission requirements,
only 2 requirements were waived. These were due to minor variances on the EFI noise floor
and clearances between Probes on the Carrier.
    Final Probe dry masses were 4.7% below their 81.8 kg Not-to-Exceed limits and matched
to less than 1%. Probes A-E measured 78.0, 77.6, 76.7, 76.7, and 78.1 kg, respectively. The
Probe Carrier Assembly weighed in at 777 kg with a 6.3% margin to the launch vehicle NTE
of 829 kg. The final Probe power budget at 38.7W has a 4.9% margin in surviving a 3-hour
eclipse at the end of mission.

7.2 Schedule Performance
    The successful scheduling of the THEMIS project required a mixture of optimism,
dogged determination and endless coordination between UCB, ATK Space and GSFC
management. From project funding to launch took 46 months, including a 4-month launch
vehicle delay. The first Probe began integrating instruments at 31 months and all Probes
completed all testing 9 months later.
    The THEMIS top-level schedule is illustrated in figure 7.2-1. The master schedule linked
together 23 instrument and 10 probe and carrier schedules, together totaling more than 5000
tasks. These were managed by schedulers at UCB and ATK Space and milestone-tracked by
    The development of instrument suites and probe buses used a 1, 2 and 2 approach to build
the five Probes. The Environmental Verification Tests (EVT) used a different approach,
verifying the first Probe then all five Probes and Carrier. The latter sequence provided “test-
as-you-fly” configurations and shortened the overall project schedule.
    All levels of integration, whether instrument or bus or probe, witnessed dramatic schedule
improvements with each unit. These were principally due to [1] improved test procedures
and [2] reduction in newly discovered problems.

                        Figure 7.2-1 THEMIS Constellation Schedule
7.3 Cost Performance
The THEMIS budgetary estimate at Confirmation Review was 158.3M (FY02) assuming a
payload development and operations of $89.3M and a launch vehicle at $69M. Actual
payload costs ended 4% high at $92.9M. Actual costs of the launch vehicle and the launch
delay brought the total cost to $172.8M, yet still beneath the MIDEX cap of $180M.

          THEMIS Budget Performance FY02 $M         Phase A-D    Phase E      Total
          UCB/SAI Probe & Carrier Development             80.3       10.3        90.6
          JPL Environments and Simulations                 1.1        0.0         1.1
          GSFC Thermal and Data Analysis                   0.4        0.9         1.2
          Payload (w/o Launch Delay)                      81.7       11.1        92.9
          Launch Vehicle (AO)                             69.0                   69.0
          Launch Vehicle Additional Costs                  4.9                    4.9
          UCB/SAI Impact of Launch Delay                   3.1         3.0        6.0
          Total                                          158.7        14.1      172.8
                           Figure 7.3-1 THEMIS Cost in FY02 M$
The project experienced dramatic cost reductions as the units were fabricated. The design
phase through CDR cost $20M, roughly one quarter the total, while costs of the first flight-
Probe were another $20M. Remarkably, the four remaining Probes were completed at
roughly half the cost of the first.

    The THEMIS mission is that rare combination of inspiration and imagination that
challenges scientists, engineers, and managers alike. Without question, the success of the
project is due to the indefatigable efforts and contagious optimism of the PI Vassilis
Angelopoulos, who not only convinced all of us that it could be done, but that we could do it.
    Dr. D. Pankow provided vehicle dynamics and flawlessly led the vibroacoustics and
balance efforts. Dr. Auslander and UCB graduate students simulated vehicle dynamic
behavior due to fuel slosh and wire booms. D. Curtis and S. Harris carefully verified each
bus as C. Chen, H. Richard and M. Ludlam did the same for the instrument suites. Dr. M.
Sholl verified the propulsion system throughout integration and led probe fueling, and M.
Leeds provided RCS training and support. Dr. M. Bester verified communications with BGS
and the Mission Ops Center.
    Deputy Project Manager D. King and Project Scheduler D. Meilhan tracked an
unbelievable number of tasks and kept it all under control as Financial Manager K. Harps
kept costs in line. Mission Assurance Manager R. Jackson and quality personnel J. Fisher
and C. Scholz managed to get all the parts, inspect all the components and track all the
problems to closure.
    ATK Space system engineers K. Brenneman and W. Chen, propulsion designer M.
McCullough, separation system designer D. Jarosz and thermal engineer R. Zara were
instrumental in the success of the probes and carrier. The Hammers company for excellent
support of Bus flight software. ATK Space Vice President F. Hornbuckle committed the
company to the project at a time when it was probably unpopular to do so.
    Mission Manager F. Snow demonstrated great patience and support particularly while the
project went through difficult times. The Explorers team of D. Lee, R. Miller, J. Thurber and
D. Gates provided support where needed in government services, scheduling, RF expertise
and communications. Finally, IIRT co-chairman Mark Goans and Brian Keegan provided
valuable reviews, personal counsel and much-appreciated confidence in the THEMIS design
and implementation.


To top